Is cardiopulmonary transit time (CPTT) measured by using dynamic rubidium cardiac PET/CT a predictor for cardiac function?

Cardiopulmonary transit time (CPTT) represents the time needed for the circulation of blood from the right to the left ventricle. This parameter can be measured during dynamic acquisition of rubidium ([⁸²Rb]) cardiac PET/CT. To further characterize this marker, we wanted to assess the association between CPTT and parameters of cardiac function derived from echocardiography. Retrospective single center analysis of patients referred to [⁸²Rb]RbCl-PET/CT with rest/stress protocol on an integrated hybrid PET/CT system (Biograph mCT, Siemens Healthineers, Erlangen, Germany) and echocardiography within 100 days. After intravenous injection of 7.5 MBq/kg [⁸²Rb]RbCl dynamic scans with initially 12 × 10 s frames were started. For data analysis a volume of interest (VOI) was drawn in the left and right ventricle using dedicated software. The difference between the peak time for the two time activity curves (TAC) was extracted as CPTT and normalized for heart rate (NCPTT). Associations between NCPTT and echo parameters such as left ventricular ejection fraction (EF_Echo) were analyzed using linear regression models. 44 patients (sex: 28 male, 16 female) were enrolled with a time difference between PET and echocardiography of 19.65 ± 23.3 days. 9 patients had a rest CPTT of 0 s, 32 patients 10 s and 3 patients 20 s. The association between EF_Echo and NCPTT revealed a significant negative correlation (beta = -0.77; CI: -1.32, -0.22; p = 0.007). Given this association, univariate predictive models for EF_Echo were applied. Root mean square error was 6.83% for the EF_PET, and 6.0% for NCPTT, which indicates a slightly higher predictive performance for the NCPTT model with a lower error. Pulmonary transit time can be estimated with [⁸²Rb]RbCl-PET/CT, with a high positive association to rest EF_Echo. However, smaller time frames than 10 s are needed, for more accurate estimation of cardiac function.